sign in sign up
<<
Home , Natural selection

Evidence for in response to natural selection in a contemporary

Emmanuel Milota,1, Francine M. Mayera, Daniel H. Nusseyb, Mireille Boisverta, Fanie Pelletierc, and Denis Réalea

aDépartement des Biologiques, Université du Québec à Montréal, Montréal, QC, Canada H3C 3P8; bInstitute of Evolutionary , University of , Edinburgh EH9 3JT, ; and cDépartement de Biologie, Université de Sherbrooke, Sherbrooke, QC, Canada J1K 2R1

Edited by Peter T. Ellison, Harvard University, Cambridge, MA, and approved August 30, 2011 (received for review March 17, 2011) It is often claimed that modern have stopped evolving However, selection measured at the phenotypic level does not because cultural and technological advancements have annihi- necessarily imply a causal relationship between the trait and fit- lated natural selection. In contrast, recent studies show that ness (12, 13) and, as a consequence, such predictions will often selection can be strong in contemporary . However, be inappropriate in the case of natural populations (14). This also detecting a response to selection is particularly challenging; pre- implies that phenotypic changes, even those occurring in the vious evidence from wild animals has been criticized for both predicted direction, may not provide robust evidence of evolu- applying anticonservative statistical tests and failing to consider tion, as they may not be indicative of underlying genetic trends random . Here we study -history variation in an (15–17). These problems are likely exacerbated in long-lived insular preindustrial French-Canadian population and apply a re- such as humans, where within- plastic responses cently proposed conservative approach to testing - to environmental variation, or viability selection, can drive phe- ary responses to selection. As reported for other such societies, notypic changes over the timescale of a study in the same di- natural selection favored an earlier age at first (AFR) rection as that predicted for genetic responses to selection (15). among women. AFR was also highly heritable and genetically To overcome these problems, recent studies of wild birds and correlated to fitness, predicting a microevolutionary change to- mammals have tested for microevolution by directly measuring ward earlier reproduction. In agreement with this prediction, AFR – declined from about 26–22 y over a 140-y period. Crucially, we changes in breeding values (16 22; see ref. 23 for a review). The uncovered a substantial change in the breeding values for this breeding value (BV) of an individual is the additive effect of his/ trait, indicating that the change in AFR largely occurred at the her on a trait value relative to the mean in the genetic level. Moreover, the genetic trend was higher than population, in other words the heritable variation that parents expected under the effect of random genetic drift alone. Our transmit to their offspring (11). In quantitative genetic (QG) results show that microevolution can be detectable over relatively notation, the phenotypic measurement can thus be written as zi = few generations in humans and underscore the need for studies of μ + ai + εi, where μ is the population average, ai is the breeding human demography and reproductive to consider the role value of individual i, and εi is a residual term that may include of evolutionary processes. environmental and nonadditive genetic effects and measurement . By definition, observing a change in BVs in the direction reproductive timing | | Homo sapiens | life-history traits | predicted by selection would constitute direct evidence for mi- lifetime reproductive croevolution. However, true BVs are not observable and must be predicted using QG models. Although a handful of studies have arwinian evolution is often perceived as a slow process. documented trends in predicted breeding values (PBVs) consis- DHowever, there is growing awareness that microevolution, tent with a microevolutionary response to selection (e.g., 19–21), defined as a genetic change from one generation to the next in it has become apparent that the statistical tests used in these response to natural selection, can lead to changes in the phe- studies were highly anticonservative (23, 24). Moreover, thus far notypes (observable characters) of over just a few studies have not excluded the possibility that observed genetic years or decades (1, 2). This likely applies to humans as well changes are similar to those expected under genetic drift, that is, because (i) natural selection operates on several morphological, the random sampling of genes between generations. physiological, and life-history traits in modern societies through It follows that empirical support for microevolution from differential reproduction or survival (3, 4), and (ii) a number of longitudinal studies of long-lived species remains sparse and these traits show heritable (4–7), attesting the controversial (15, 23). Here we investigate the genetic basis of potential for a microevolutionary response to selection. This age at first reproduction (AFR), a good candidate for an evolving evolutionary potential of modern humans has major implica- trait in humans (4). We used a recently advocated Bayesian tions. First, it signifies that we should consider the role of evo- quantitative genetic approach (23) to test whether advancement lutionary processes that might underlie any observed trends in in women’s AFR that occurred over a 140-y period in a French- . Second, it may produce eco-evolutionary feedbacks Canadian preindustrial population was attributable to micro- modifying the dynamics of modern populations (2, 8). This also evolution. We uncovered a genetic response to selection in this means that the accuracy of forecasts, for instance those per- key life-history trait, with potentially important demographic taining to demography or , and on which public consequences for this population. policies may rely, could well depend on our knowledge of contemporary evolution.

However, identifying which traits are evolving in which pop- Author contributions: E.M., F.M.M., and D.R. designed research; E.M., M.B., and F.M.M. ulation is technically difficult. First, it requires information on performed research; E.M., D.H.N., F.P., and D.R. analyzed data; and E.M., F.M.M., D.H.N., phenotype, pedigree links, and fitness over a sufficient number of M.B., F.P., and D.R. wrote the paper. generations (9), which is rarely available. Second, robustly The authors declare no conflict of interest. demonstrating a response to selection is challenging. Typically, This article is a PNAS Direct Submission. phenotypic trends observed in populations are compared with 1To whom correspondence should be addressed. E-mail: [email protected]. evolutionary predictions based on selection and heritability This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. estimates, for example, using the ’s equation (10, 11). 1073/pnas.1104210108/-/DCSupplemental.

17040–17045 | PNAS | October 11, 2011 | vol. 108 | no. 41 www.pnas.org/cgi/doi/10.1073/pnas.1104210108 Downloaded by guest on September 29, 2021 Population of Ile aux Coudres associated with fitness through fertility (direct standardized selec- Ile aux Coudres is a 34-km2 island located ∼80 km to the tion gradient: −0.486; Table S1). There was also a positive asso- northeast of Québec City along the St. Lawrence River (Can- ciation between age at last reproduction (ALR) and LRS (again ada). Thirty families settled on the island between 1720 and 1773 through fertility), indicating a fitness advantage to women with and the population reached 1,585 people by the 1950s (25) (Fig. longer reproductive lifespan (Fig. 1). However, the existence of an S1). This population is ideal to study the genetic basis of life- evolutionary tradeoff between reproduction and maintenance history traits (LHTs) (Table 1). First, church registers provide functions (35) is suggested by the positive phenotypic correlation exceptionally detailed records of dates of births, marriages, and between AFR and ALR (Fig. 1), meaning that women who began . Second, the long-term data and endogamy (marriages reproducing at a younger age also tended to stop at a younger age. within the population) provide a deep and intricate pedigree to As a result, selection on one trait was counterbalanced by selection facilitate the separation of genetic and environmental influences on the other trait (Table S1). Marriage–first birth interval (MFBI), on LHTs (26). Third, the population was very homogeneous used as a proxy for fecundity (capacity to conceive; Materials and among families, particularly in traits known to correlate with the Methods), had a significant direct effect on AFR (Fig. 1), sug- timing of reproduction (social class, education, and ) (3, gesting that the variation in AFR is partly due to variation in fe- 27). In addition, the split of resources among families was quite cundity among women (or couples). However, MFBI was very even due to the type of land distribution, and the number of weakly and not significantly correlated to fertility, suggesting that professions was limited (SI Text 1). This relative homogeneity the reproductive lifespan has a greater influence on fertility than should minimize confounding socioeconomic or shared envi- fecundity per se, or that factors other than fecundity (e.g., lactation ronmental influences within quantitative genetic analyses. amenorrhea) (36) had an important influence on the reproductive We examined the life history of women married after 1799, as rates beyond the first child. Finally, longevity had a small direct the genealogical depth is highest after this date, and before 1940, effect on fitness but was under strong indirect and positive selection to make sure that the couples retained had completed their owingtoitsstrongcorrelationwithALR(Fig.1;Table S1). family before the records ended (in 1973). Following ref. 28, we AFR was significantly heritable, predicting a microevolution- used two different datasets that make different assumptions re- ary change toward earlier first reproduction given that the trait is garding unusually long interbirth intervals in the demographic under . We used a Bayesian implementation records. The “subfecundity” dataset (n = 572 women) assumes (37) of linear mixed-effects animal models (26) to estimate the

that unusually long interbirth intervals reflect subfecundity. The heritability in AFR and LRS while controlling for the effects of EVOLUTION “migration” dataset (n = 363 women) assumes that long inter- shared familial environment, , temporal trends, and vals may also reflect emigration from the island and excludes whether a woman gave birth to twins (Materials and Methods). families with such length intervals (see SI Text 2 for data-filtering Heritability was high for AFR (0.30 and 0.55, depending on the criteria and Table 1 for average life-history trait values). dataset used) and low for LRS (<0.01 and 0.04; Table 2). The presence of a strong negative genetic correlation between AFR Selection on Age at First Reproduction and LRS (Table 2) further supports the potential for a genetic The adaptive significance of the timing of reproduction is well- response to selection (14), although some uncertainty is associ- established within (29), including in humans ated with this correlation resulting from uncertainty in estimates (30). In particular, selection in favor of earlier AFR has been of the heritability in LRS in our models (Materials and Methods). ANTHROPOLOGY previously documented in several pre- and postindustrial human The shared familial environment had a negligible effect on both societies (3, 4, 7, 27, 31). French-Canadian preindustrial societies traits (Table 2). exhibited a natural fertility, that is, non-Malthusian, regime (32). In the absence of birth control methods, the full reproductive Genetic Response to Selection potential of couples can be expressed. Consequently, earlier re- Average AFR advanced from about 26 to 22 y over the study production may lead to bigger family size and confer higher fit- period (Fig. 2), therefore in the direction predicted by selection. ness, in particular at time of population expansion (33), provided We tested for a genetic response to selection by comparing that fertility correlates with fitness (SI Text 1). temporal trends in the breeding values predicted by our Bayesian On île aux Coudres, selection indeed strongly favored women models (PBVs) with trends in breeding values randomly gener- with earlier AFR. A path analysis (34) accounting for selection ated along the pedigree under a scenario of pure random genetic on other life-history traits correlated to AFR showed a negative drift (RBVs) (23). We found a negative trend in PBVs that was association between AFR and fertility (completed family size), steeper than expected under drift alone (Fig. 2). Remarkably, the whereas fertility is itself strongly associated with lifetime re- estimated genetic change in AFR corresponded to a decline of productive success [LRS; used as a proxy for fitness (4)] [results up to 3 y between the first and last cohorts (Table 2), thus for the subfecundity dataset in Fig. 1 and Table S1; the migration explaining a substantial part of the observed phenotypic change dataset led to similar results (Fig. S2)]. Therefore, AFR is negatively between 1800 and 1939.

Table 1. Average phenotypic values (±SD) for female life-history traits in the preindustrial human population of île aux Coudres Women included under the Trait Migration dataset* Subfecundity dataset subfecundity hypothesis only

Marriage–first birth interval (mo) 13.9 ± 6.2 (360) 17.8 ± 22.0 (564) 25.7 ± 34.6 (204) Age at first birth (y) 23.4 ± 3.9 (363) 23.8 ± 4.3 (572) 24.5 ± 4.9 (209) Age at last birth (y) 38.7 ± 6.7 (363) 36.1 ± 7.3 (572) 31.6 ± 6.1 (209) Longevity (y) 56.9 ± 22.2 (252) 58.2 ± 21.6 (301) 65.1 ± 17.0 (49) Fertility (completed family size) 8.6 ± 3.9 (363) 7.0 ± 4.1 (572) 4.3 ± 2.9 (209) Lifetime (offspring 7.0 ± 3.4 (363) 5.1 ± 3.5 (363) 3.5 ± 2.6 (209) living to age 15)

Sample size is in parentheses. *See SI Text 2 for dataset description.

Milot et al. PNAS | October 11, 2011 | vol. 108 | no. 41 | 17041 Downloaded by guest on September 29, 2021 0.93 0.30 (0.08)8)) (0.06) -0.01 (0.03)

0.20 -0.51 0.25 (0.06) (0.03) (0.02)(

0.15 0.74 0.960 (0.06) (0.04) (0.02)

0.67 0.06 0.06006 (0.07) (0.04) 0.02 (0.01)

(0.02)

Fig. 1. Path diagram describing the selection exerted on female life-history traits at île aux Coudres. Solid one-way arrows show presumed causal rela- tionships between variables, and dashed two-way arrows are noncausal correlations. Values (±SEM) next to solid arrows are standardized regression coef- ficients (direct effects for selection gradients), and values next to dashed arrows are correlation coefficients. Values (±SEM) and arrows in gray are for unmeasured causes (residual variance) of endogenous variables. Direct paths are those passing through causal relationships only (e.g., AFR > fertility > LRS), whereas indirect paths pass through at least one correlational relationship (e.g., AFR <> ALR > fertility > LRS). Life-history traits are: AFR, age of the woman at first reproduction; ALR, age of the woman at last reproduction; fertility, completed family size; longevity, woman’s lifespan; LRS, lifetime reproductive success; MFBI, marriage–first birth interval. Results are for the subfecundity dataset (n = 283; Materials and Methods); the migration dataset led to similar path coefficients (Fig. S2).

Lifetime reproductive success showed a phenotypic increase reduces the power to measure heritability and detect a trend (SI by three to four children over the study period (i.e., from 4.7 to Text 2). 7.9 children for the subfecundity dataset, and from 6.3 to 10.6 for the migration dataset; Fig. 2). Moreover, the trend in the PBVs Discussion of LRS was positive and steeper than expected by drift, sug- Throughout the history of île aux Coudres, there was a pro- gesting a temporal increase in fitness under the effect of selec- gressive advancement of age at first reproduction: Women giving tion on AFR (Fig. 2). birth to their first child around the 1930s were about 4 y younger The difference between the slopes in PBVs and RBVs was than those who began to reproduce around 1800. There was a significant in the subfecundity dataset for both AFR and LRS (P < concomitant increase in lifetime reproductive success as women 0.01; Table 2). Using the migration dataset, the difference was who began their reproduction earlier generally had more chil- nearly significant for AFR (P = 0.058) and the strong genetic dren surviving to adulthood. Whereas little information on AFR trend in PBVs was quite robust to modifications of the model is reported for other Québec populations, the age at marriage of settings or Bayesian priors (Materials and Methods). However, the women apparently remained stable in the countryside and in- difference was not significant for LRS. Differences between the creased in urbanized areas (38). AFR likely followed the same two datasets are likely to be due to the fact that, by definition, historical pattern because it should correlate positively with age the migration dataset excludes a part of the natural life-history at marriage when marriage marks the onset of reproduction. On variation of the population (particularly in LRS), which likely île aux Coudres, both traits were strongly correlated (sub-

Table 2. Genetic parameters and response to natural selection in woman’s age at first reproduction and lifetime reproductive success at île aux Coudres between 1800 and 1939 Shared familial environment Genetic correlation Heritability effects between AFR and LRS Genetic response

Dataset Response variable Mode Interval Mode Interval Mode Interval Trend PBVs Prob. drift ≥ obs.

Subfecundity AFR 0.55 0.30–0.90 0.01 0.00–0.15 −0.81 −0.97 to −0.48 −2.95 0.009 LRS 0.04 0.00–0.43 0.00 0.00–0.07 —— +0.28 0.009 Migration AFR 0.30 0.08–0.73 0.01 0.00–0.12 −0.81 −0.99 to 0.16 −1.74 0.058 LRS <0.01 0.00–0.12 0.00 0.00–0.02 —— +0.08 0.144

For heritability, shared familial environment effects, and genetic correlation, the mode of the posterior distribution (i.e., the point estimate ofthe parameter) and the 95% Bayesian posterior interval of highest density are reported separately for each dataset. The genetic correlation involves both traits and is only shown once for each dataset. The genetic response is the difference in mean PBVs between the first and last women’s birth cohorts computed from the slope of the regression of PBVs on eight 20-y cohorts (means are over all women of a cohort and 1,000 MCMC samples). The trend in PBVs is in years for AFR and on the latent scale (Poisson model) for LRS. “Prob. drift ≥ obs” indicates the of observing a trend as strong or stronger due to random genetic drift alone (two-tailed test).

17042 | www.pnas.org/cgi/doi/10.1073/pnas.1104210108 Milot et al. Downloaded by guest on September 29, 2021 EVOLUTION

Fig. 2. Temporal trends in the phenotypic and breeding values of woman’s age at first reproduction and lifetime reproductive success in the population of île aux Coudres between 1800 and 1939. All values are in years for AFR. For LRS, phenotypic values are in numbers of offspring reaching age 15, whereas PBVs are on the latent scale (Poisson model). PBVs are genotypic deviations from the population average over the study period [zero values correspond to no de- ANTHROPOLOGY viation; diamonds are averages from 1,000 MCMC samples (±SD)]. The genetic trend expected under random genetic drift alone (i.e., in randomly generated breeding values) is also shown by a dashed line. For the sake of visual comparison of slopes, the intercept of the drift trend was set to the same value as the intercept for the observed trend.

fecundity dataset: r = 0.90 [95% confidence interval (CI): 0.88– (24) and for shared familial environment effects that could bias 0.91]; migration dataset: r = 0.98 [CI: 0.98–0.99]). Moreover, the heritability estimates. Actually, there is accumulating evidence trend in LRS is associated with an increase in fertility, that is, that PBVs measured from such multigenerational pedigrees are completed family size (Fig. S3), which is also at odds with what is measuring genetic effects (e.g., 43). generally reported for Québec, especially in the first half of the 20th century (39, 40). Consequently, the trends in LHTs at île Nongenetic Hypotheses for Life-History Trends. Although the trend aux Coudres suggest that factors operated on the island in op- in breeding values we observed is consistent with a microevolu- position to socioeconomic or cultural trends operational at a tionary response to natural selection, other factors could nev- larger scale (39). Indeed, our results provide evidence that those ertheless have contributed to the temporal trends in AFR and changes resulted, at least partly, from a microevolutionary re- LRS. Most importantly, the advancement of age at maturity, as well as increases in fertility, may reflect plastic responses to sponse to natural selection on AFR. improvements in nutritional conditions, such as those observed Crucially, the above conclusion relies on the reliability of during the 19th and 20th centuries in Western societies. Better- PBVs. Here we used a Bayesian analysis intended to avoid the fed women grow faster, mature earlier and in a better physio- anticonservatism characterizing previous tests of microevolution logical state, and are more fecund (44). Importantly, alongside (23, 24). One potential issue with this approach is its sensitivity in such plastic responses in reproductive traits, we would expect an the choice of prior distributions for variance parameters (41). increase in infant and juvenile survival rates with time (45). However, the test of microevolution in AFR was robust for Despite some fluctuations, infant and juvenile survival rates on various weakly to moderately informative priors. Another po- île aux Coudres were not higher at the end of the study period tential problem is that when limited information from relatives is than at the beginning (Fig. S4). Furthermore, there is no evi- available or when relatives share similar environments, PBVs can dence that the population underwent a grasp part of the variation due to nongenetic sources (24, 42). of the sort observed elsewhere during the 19th and 20th centu- However, the animal model is robust to this kind of bias when ries. This would involve a decline in fertility and mortality supplied with deep and intricate pedigrees because it uses all alongside increasing urbanization, none of which occurred on île degrees of relatedness among to estimate genetic aux Coudres (Figs. S3 and S4; SI Text 1). Therefore, there is parameters. In addition, nongenetic sources of variation can be limited support for the that reproductive plasticity in re- accounted for explicitly. Here we controlled for temporal trends sponse to changing conditions can explain the trends in LHTs in traits that might arise from other causes than a change in BVs we observed.

Milot et al. PNAS | October 11, 2011 | vol. 108 | no. 41 | 17043 Downloaded by guest on September 29, 2021 Whereas a vast majority of men were farmers before 1870, a Phenotypic Selection Analysis. We fitted univariate general linear models diversification of occupations after that date progressively in- (GLMs) for women’s fertility (completed family size) and LRS to control for fl creased the of the island (SI Text 1). If it also temporal uctuations and other sources of variation based on preliminary analyses of the data. We thus controlled for year of marriage, whether or not meant more available per family, it perhaps contributed a couple gave birth to twins, and infant mortality (0–1 y). Inbreeding is a to the rise in fertility. However, we have no clear indication from structural characteristic of the population of île aux Coudres (51) and shows the literature that this was the case. In addition, when consid- complex relationships with LHTs (28, 52). Therefore, we also included linear and ering couples married before and after 1870 separately, selection quadratic terms of kinship between spouses (i.e., the inbreeding coefficient of gradients on AFR, ALR, and fertility were in the same direction their children). We also controlled for the common familial environment shared and of similar magnitude for the two periods (Table S2), in- by sisters (random effect) but dropped this term because of its small and non- fi dicating no substantial change in the selective regime after 1870. signi cant effect. The analysis was conducted on women for which longevity was known and data were available for all other traits (subfecundity dataset: Reproductive compensation by inbred couples, which were hy- n = 283; migration dataset: n =251;SI Text 2). We used the residuals of fertility pothetically exposed to higher infant mortality, could have in- and LRS from the GLMs in a path analysis (34) of phenotypic selection on cor- creased fertility rates (39) (note that we control for infant related traits (53) using LRS as a fitness proxy (an analysis on raw data instead mortality in our selection analyses), but evidence for this hy- gave very similar results but yielded models with slightly poorer fit; hence, we pothesis is inconclusive (28). Wealth transmission patterns pos- only report the results for the analysis on residuals). We conducted the analysis sibly contributed to create within-family variation in life history using the SEM package for R (54) and the path model described next. fi (SI Text 1). However, this alone would not explain how a non- We built a modi ed version of a path diagram of causal relationships be- tween female life-history traits and fitness that was applied by Pettay et al. (3) genetic effect could be strong enough to mimic a high heritability to a Finnish population. In this model (Fig. 1), AFR, ALR, and longevity have without being detectable as phenotypic resemblance among full direct effects on fertility and an indirect effect on LRS through fertility. Lon- sibs. Finally, cultural transmission of fitness (CTF) can cause gevity also has a direct effect on fitness because it may affect the duration of nongenetic inheritance in human traits, and was documented in parental care, and thus offspring survival. AFR, ALR, and/or longevity are the nearby Saguenay-Lac-St-Jean French-Canadian population expected to be correlated (35), and thus these correlations were included in the ’ (46). However, we would have expected CTF to be partly path diagram. One distinction with Pettay et al. s original model is the exclusion fl of the proportion of surviving offspring, because its effect should be mainly re ected in family effects, which again were negligible in all of mediated through interbirth intervals. Mean interbirth interval (MIBI) is itself our analyses. the product of other traits already included in the model: MIBI = (ALR – AFR)/ fertility. Another distinction with Pettay et al.’s model is the inclusion of the Life-History Evolution in Modern Humans. Very few empirical MFBI as a trait correlated to fertility (i.e., noncausal). The rationale is that MFBI investigations of secular changes in life-history traits in humans reflects fecundability to some degree (i.e., the probability of conceiving in have considered microevolutionary hypotheses. Certainly, these a given month) (39), as opposed to interbirth intervals, which also depend on should not be discarded a priori simply because an immediate lactation amenorrhea (36) and perhaps on care demands by older children. In nongenetic explanation may exist. In particular, natural selection turn, fecundability should be tightly related to fecundity, the physiological ca- pacity to conceive. Consequently, MFBI is perhaps the best proxy that we have on reproductive timing appears to be widespread in humans, for fecundity for the île aux Coudres population (i.e., MFBI should decrease with whereas AFR was found to be heritable in several contemporary increasing fecundity). populations, with an across-study average of 0.11 (4). Moreover, at least one other study uncovered a negative genetic Estimation of Genetic Parameters. We fitted bivariate “animal” models (26), a between AFR and LRS [in an American population (7)], which is type of generalized linear mixed-effects model (GLMM), to estimate the ad- a better predictor of the response to selection than the breeder’s ditive genetic variance (Va) of AFR and LRS and their genetic correlation, as well equation (14). Clearly, the potential for genetic responses of the as the breeding values for each woman. The animal model uses the in- kind observed here is not just limited to the île aux Coudres formation from all pedigree relationships to specify the expected phenotypic resemblance between relatives. It has several advantages for the study of wild population. However, only through the wider application of the populations, including its power to separate environmental from genetic approaches used here to other human populations can we estab- sources of resemblance between relatives (especially with an intricate pedigree lish their generality. structure), its applicability to unbalanced sampling designs, and its Our study, as well as previous investigations, raises the ques- to departures from distributional assumptions (11, 26). The Bayesian imple- tion of why a trait like AFR would be heritable. Actually, heri- mentation of GLMMs in the MCMCglmm R package (37) was used to fitmodels table traits such as growth rate and birth weight likely correlate independently for the subfecundity and migration datasets. Again, we con- trolled for temporal trends of environmental origin by entering the year of positively with age at maturity in humans (44, 45). Age at men- marriage (24) and for inbreeding (quadratic effect). Whether a woman gave arche could play a pivotal role here, as it also correlates with birth to at least one pair of twins was found to affect LRS in the above GLMs, these traits on the one hand (e.g., 47) and with both age at and hence this factor was entered in the LRS models. We controlled for the

marriage and AFR in human societies with drastically different familial environment shared by sisters (VCE) by entering the marriage identi- cultures (48). Incidentally, age at menarche was repeatedly fication of the woman’s parents (here confounded with maternal effects be- found to be heritable (typical heritability around 0.5) (49). cause only full sibs are known in this population). The distribution of AFR was Our study supports the idea that humans are still evolving. It modeled as Gaussian and that of LRS as Poisson. Samples were taken from the posterior distributions of Va, VCE, and the residual variance (Vr) every 7,500 also demonstrates that microevolution is detectable over just a iterations of the Markov chain after an initial burn-in of 1,500,000 iterations, few generations in long-lived species. For instance, a large pro- for a total of 1,000 samples. For each Markov chain Monte Carlo (MCMC) portion of the phenotypic trend in age at first reproduction at île sample from bivariate models, the narrow-sense heritability (h2) of AFR was aux Coudres appears to be attributable to a response to natural calculated as Va/Vp,whereVp = Va + VCE + Vr is the phenotypic variance, 2 β selection. Modifications in the timing of reproduction can have whereas h of LRS was calculated on the latent scale as Va/(Vp + ln(1/exp( 0)+1)), β important effects on the demography of a population (e.g., 50). where 0 is the intercept of the Poisson model (55). The shared familial envi- Therefore, human studies need to carefully consider the role of ronment effects were calculated likewise, except that Va was replaced by VCE in the numerator. The genetic is reported here in the stan- microevolutionary processes underlying any observed trends in 2 dardized form of the genetic correlation (rG). The posterior mode of h and rG traits and their potential feedback on . was used as point estimates, whereas Bayesian 95% intervals of highest density were used to test whether these estimates differed significantly from zero. Materials and Methods Lifetime Reproductive Success. We calculated the LRS of a woman as the Testing for an Evolutionary Response to Selection. We used a method recently number of her children who survived to age 15 y old, that is, approximately the advocated by Hadfield et al. (23) to test for a response to selection while

minimal age at marriage at île aux Coudres (see SI Text 2 for further details). accounting for drift: the posterior estimate of Va from a given MCMC

17044 | www.pnas.org/cgi/doi/10.1073/pnas.1104210108 Milot et al. Downloaded by guest on September 29, 2021 sample from the bivariate model of AFR and LRS fitted above was used to trend in PBVs of LRS was always higher than expected by drift but not always randomly generate breeding values along the pedigree of île aux Coudres significantly so. This greater fluctuation of LRS with prior choice is likely under a scenario of pure random genetic drift (RBVs), using the rbv explained by the fact that the heritability of LRS is low and because Bayesian of the MCMCglmm package. Then, mean RBVs were regressed parameter estimation is more difficult in those cases. against cohort (eight 20-y cohorts), and the slope coefficient (βRBV)was compared with that (β ) of the regression of PBVs against the cohort for PBV ACKNOWLEDGMENTS. We thank Jarrod Hadfield and Bill Shipley for the same MCMC sample. This procedure was repeated for all MCMC statistical advice, and Jérôme Laroche, Stéphane Larose, and the Centre de β samples. The proportion of times where the absolute value of RBV was as Bioinformatique (Université Laval) for computing support. This project was β high or higher than the absolute value of PBV was taken as the probability funded by the Fonds Québécois de la Recherche sur la et les Technol- of obtaining the observed genetic trend (i.e., in PBVs) as the result of drift ogies (E.M.) and the Canada Research Chair in Behavioural Ecology (D.R.). only (i.e., two-tailed test). D.H.N. was supported by a Research Council postdoc- toral fellowship and a Biotechnology and Biological Sciences Research Council Bayesian Prior Choice and Testing. Several priors were tested to finally retain David Phillips Fellowship. Pierre Philippe originally built the île aux Coudres the least informative ones leading to proper posterior distributions for database in 1967 with Jacques Gomila, Jean Benoist, and Guy Dubreuil (Uni- variance parameters in the Bayesian models. Thus, in bivariate models, we versité de Montréal) and the financial support of the Canada Council for the used moderately informative priors: Variance parameters (V) were set to 1 Arts. Since 1986, the register was computerized and updated by F.M.M., M.B., (and to zero) and the degree of belief (nu) to 2. We also ran Yolande Lavoie, and Pierre Philippe, successively with the financial support of univariate models with various weakly informative priors (e.g., V =1,nu = the Université de Montréal, the Fonds pour la Formation de Chercheurs et 0.002). The trend in PBVs of AFR was robust and significantly higher than l’Aide à la Recherche du Québec, and the Social Sciences and Humanities Re- drift whatever the priors used in uni- or bivariate models (except for the mi- search Council of Canada. Since 1988, the database was integrated and man- gration model in Table 2, where the trend is close to significance: P =0.058).The aged in the ANALYPOP software developed in F.M.M.’s laboratory.

1. Hairston NGJ, Ellner SP, Geber MA, Yoshida T, Fox JA (2005) Rapid evolution and the 28. Boisvert M, Mayer FM (1994) Infant mortality and consanguinity in an endogamous convergence of ecological and evolutionary time. Ecol Lett 8:1114–1127. population in Québec. Population (Paris), 49:685–724 (in French). 2. Carroll SP, Hendry AP, Reznick DN, Fox CW (2007) Evolution on ecological time-scales. 29. Stearns SC (1992) The Evolution of Life Histories (Oxford Univ Press, Oxford). Funct Ecol 21:387–393. 30. Hawkes K, Paine RR (2006) The Evolution of Human Life History (School Am Res Press, 3. Pettay JE, Helle S, Jokela J, Lummaa V (2007) Natural selection on female life-history Santa Fe, NM). traits in relation to socio-economic class in pre-industrial human populations. PLoS 31. Käär P, Jokela J, Helle T, Kojola I (1996) Direct and correlative phenotypic selection on One 2:e606. life-history traits in three pre-industrial human populations. Proc Biol Sci 263: 4. Stearns SC, Byars SG, Govindaraju DR, Ewbank D (2010) Measuring selection in con- 1475–1480. temporary human populations. Nat Rev Genet 11:611–622. 32. Henripin J (1957) From acceptance of nature to control: The demography of the EVOLUTION 5. Pettay JE, Kruuk LEB, Jokela J, Lummaa V (2005) Heritability and genetic constraints French Canadians since the seventeenth century. Can J Econ Polit Sci 23:10–19. of life-history trait evolution in preindustrial humans. Proc Natl Acad Sci USA 102: 33. Brommer JE, Merilä J, Kokko H (2002) Reproductive timing and individual fitness. Ecol 2838–2843. Lett 5:802–810. 6. Kosova G, Abney M, Ober C (2010) Colloquium papers: Heritability of reproductive 34. Scheiner SM, Mitchell RJ, Callahan HS (2000) Using path analysis to measure natural fitness traits in a human population. Proc Natl Acad Sci USA 107(Suppl 1):1772–1778. selection. J Evol Biol 13:423–433. 7. Byars SG, Ewbank D, Govindaraju DR, Stearns SC (2010) Colloquium papers: Natural 35. Kirkwood TBL, Austad SN (2000) Why do we age? Nature 408:233–238. selection in a contemporary human population. Proc Natl Acad Sci USA 107(Suppl 1): 36. Philippe P (1974) Amenorrhea, intrauterine mortality and parental consanguinity in 1787–1792. an isolated French Canadian population. Hum Biol 46:405–424. 8. Pelletier F, Garant D, Hendry AP (2009) Eco-evolutionary dynamics. Philos Trans R Soc 37. Hadfield JD (2010) MCMC methods for multi-response generalised linear mixed –

Lond B Biol Sci 364:1483 1489. models: The MCMCglmm R package. J Stat Softw 33, issue 2. ANTHROPOLOGY 9. Pemberton JM (2008) Wild pedigrees: The way forward. Proc Biol Sci 275:613–621. 38. Charbonneau H (1991) The Population of Québec from Yesterday to Tomorrow, eds 10. Falconer DS, Mackey TFC (1996) Introduction to Quantitative (Longmans Henripin J, Martin Y (Presses de l’Université de Montréal, Montréal), pp 11–23 (in Green, Harlow, UK). French). 11. Lynch M, Walsh B (1998) Genetics and Analysis of Quantitative Traits (Sinauer, Sun- 39. Philippe P (1973) Fertility, fecundability and consanguinity at Isle-aux-Coudres. Re- derland, MA). cherches Sociobiographiques 14:117–123 (in French). 12. Wade MJ, Kalisz S (1990) The causes of natural selection. Evolution 44:1947–1955. 40. Henripin J (1991) The Population of Québec from Yesterday to Tomorrow, eds 13. Rausher MD (1992) The measurement of selection on quantitative traits: Biases due to Henripin J, Martin Y (Presses de l’Université de Montréal, Montréal), pp 45–50 (in environmental covariances between traits and fitness. Evolution 46:616–626. French). 14. Morrissey MB, Kruuk LEB, Wilson AJ (2010) The danger of applying the breeder’s 41. Gelman A (2006) Prior distributions for variance parameters in hierarchical models. equation in observational studies of natural populations. J Evol Biol 23:2277–2288. Bayesian Anal 1:515–533. 15. Gienapp P, Teplitsky C, Alho JS, Mills JA, Merilä J (2008) Climate change and evolu- 42. Mrode RA (1996) Linear Models for the Prediction of Values (CABI, tion: Disentangling environmental and genetic responses. Mol Ecol 17:167–178. Wallingford, UK). 16. Merilä J, Kruuk LEB, Sheldon BC (2001) Cryptic evolution in a wild bird population. 43. Tomkins JL, Penrose MA, Greeff J, LeBas NR (2010) Additive genetic breeding values Nature 412:76–79. correlate with the load of partially deleterious . 328:892–894. 17. Wilson AJ, et al. (2007) of growth and cryptic evolution of body 44. Frisch RE (1978) Population, intake, and fertility. There is historical evidence for size in an island population. Evol Ecol 21:337–356. a direct effect of on reproductive ability. Science 199:22–30. 18. Kruuk EB, et al. (2002) Antler size in red : Heritability and selection but no 45. Stearn SC, Koella JC (1986) The evolution of in life-history traits: evolution. Evolution 56:1683–1695. Predictions of reaction norms for age and size at maturity. Evolution 40:893–913. 19. Coltman DW, et al. (2003) Undesirable evolutionary consequences of trophy hunting. 46. Heyer E, Sibert A, Austerlitz F (2005) Cultural transmission of fitness: Genes take the Nature 426:655–658. fast lane. Trends Genet 21:234–239. 20. Réale D, McAdam AG, Boutin S, Berteaux D (2003) Genetic and plastic responses of 47. Terry MB, Ferris JS, Tehranifar P, Wei Y, Flom JD (2009) Birth weight, postnatal a northern mammal to climate change. Proc Biol Sci 270:591–596. growth, and age at menarche. Am J Epidemiol 170:72–79. 21. Garant D, Kruuk LEB, Wilkin TA, McCleery RH, Sheldon BC (2005) Evolution driven by 48. Udry JR, Cliquet RL (1982) A cross-cultural examination of the relationship between differential dispersal within a wild bird population. Nature 433:60–65. ages at menarche, marriage, and first birth. Demography 19:53–63. 22. Teplitsky C, Mills JA, Alho JS, Yarrall JW, Merilä J (2008) Bergmann’s rule and climate 49. Towne B, et al. (2005) Heritability of age at menarche in girls from the Fels Longi- change revisited: Disentangling environmental and genetic responses in a wild bird tudinal Study. Am J Phys Anthropol 128:210–219. population. Proc Natl Acad Sci USA 105:13492–13496. 50. Goldstein JR, Schlag W (1999) Longer life and . Popul Dev Rev 25: 23. Hadfield JD, Wilson AJ, Garant D, Sheldon BC, Kruuk LEB (2010) The misuse of BLUP in 741–747. ecology and evolution. Am Nat 175:116–125. 51. Philippe P, Gomila J (1972) Inbreeding effects in a French Canadian isolate. I. Evolu- 24. Postma E (2006) Implications of the difference between true and predicted breeding tion of inbreeding. Z Morphol Anthropol 64:54–59. values for the study of natural selection and micro-evolution. J Evol Biol 19:309–320. 52. Phillippe P (1977) Genetics of fecundity: A demographic approach. Hum Biol 49: 25. Martin Y (1957) Île-aux-Coudres: Population and economy. Cahiers de Géographie,2: 11–18. 167–195 (in French). 53. Lande R, Arnold SJ (1983) The measurement of selection on correlated characters. 26. Kruuk LEB (2004) Estimating genetic parameters in natural populations using the Evolution 37:1210–1226. ‘animal model.’ Philos Trans R Soc London Ser B 359:873–890. 54. Fox J (2010) Package SEM (R Project; http://www.r-project.org), 0.9-21. 27. Kirk KM, et al. (2001) Natural selection and quantitative genetics of life-history traits 55. Nakagawa S, Schielzeth H (2010) Repeatability for Gaussian and non-Gaussian data: A in Western women: A twin study. Evolution 55:423–435. practical guide for . Biol Rev Camb Philos Soc 85:935–956.

Milot et al. PNAS | October 11, 2011 | vol. 108 | no. 41 | 17045 Downloaded by guest on September 29, 2021